Nate Silver's 'On The Edge' Is Meh But Read It Anyway
Book notes on a disappointing read
I really enjoyed The Signal and the Noise and like his blog so mismatched expectations were my downfall.
Why Read It At All? It will be popular among the podcast circuit and thinkbois so it will help you have pseudo-intellectual conversations on Zoom calls and networking events. Kinda like what discussing Sapiens was a few years ago.
The book lacks a unifying concept/theory and is something between a memoir and gonzo journalism. There is no big idea, it’s a interview-driven tour of how people of the River (Silicon Valley, Wall Street, gamblers) use probability, and their relationship to The Village (the Liberal Establishment).
There are other communities in the River, though: Silicon Valley, Wall Street, sportsbetting, crypto, even effective altruism, all of which are covered extensively in the book. And I found I had a lot in common with these people too, even if I sometimes disagree with their politics. There are traits like decoupling, contrarianism and a high risk tolerance that I share with the River, for better or worse. And these seem to be correlated with extremely high-variance outcomes: tremendous success or tremendous failure (as in the case of Sam Bankman-Fried, who is sort of the antihero of the book).
The rival community to the River is what I call the Village. This term is more familiar: it’s basically the liberal establishment. Harvard and the New York Times; academia, media and government. The Village is politically progressive and, I would argue, increasingly partisan — though it can also be highly competent and it’s pretty good at winning elections.
He then tours the different parts of The River: Upriver (thought leaders of risk, utilitarians, Effective Altruists, etc.), Midriver (Wall Street, Silicon Valley, etc.), Downriver (gamblers), and The Village (basically, Curtis Yarvins’s ‘the cathedral‘).
Silver’s writes about his poker days, and describes the economics and life of being a pro poker player with some colorful characters from that circuit, how they perceive & manage risk, how they came to be professional risk takers, what makes them tick, and he also describes the business models of casinos, and a lot insider baseball of sports betting. I enjoyed this part and wish he would write more about this, and I particularly liked this graphic.
The Gell-Mann Amnesia Effect
The phenomenon of a person trusting newspapers for topics which that person is not knowledgeable about, despite recognizing the newspaper as being extremely inaccurate on certain topics which that person is knowledgeable about.
He made elementary mistakes for a guy who claims to know a lot about markets and many Wall Street people. For example, he says that "hedge funds try to beat the index funds". Hedge funds do not do this nor aim to do this. Hedge funds aim give their investors no- or low-net exposure to markets, i.e. uncorrelated returns. The hedge fund return stream is supposed to be closer to an absolute return-esque product that complements the rest of their portfolio. They are stabilizers or ballast for when the rest of your portfolio’s assets are volatile.
Another mistake: Silver claims that the Thiel fellowship "spawned founders like Vitalik Buterin" and the $100k grant enabled him to drop out but Buterin had already dropped out and created Ethereum before "he was invited to apply" for the Thiel Fellowship. A charitable interpretation of this is that he was fooled by the "logo/founder shopping" by people who run the Thiel Fellowship.
The Gell-Mann Amnesia effect meant that I discounted most other sections of the book.
An important concept of risk I would have expected to find in the book but didn’t would have been Sequence Risk though there is a little about avoiding the Risk of Ruin/blowing up/Ergodicity, in the context of Sam Bankman-Fried being poor at understanding this.
SF-Bay Area and Venture Capital
Silver writes about VCs having to be slightly ahead of the curve/investing ahead of the consensus, i.e. recursively investing (I have written about Recursive Investing earlier) because their markups happen only when other VC follow them to invest in their portfolio companies. He uses the Keynesian Beauty Contest thought experiment to describe this though he updates the thought experiment from having a newspaper competition to a Buzzfeed competition, and avoids using a female beauty pageant to make his point.
He points to failed coordination/collective fallacy as roughly a cause of why black, Hispanic, and female entrepreneurs get a low share of VC funding. There is an Abilene Paradox situation: while an individual VC is not racist and sexist but this particular VC thinks that all their VC colleagues are racist and sexist so the black, Hispanic, and female founders will have a hard time raising their next round of funding and the said VC will not be able to markup their portfolio and hence make it harder to raise their next fund to grow their assets under management & management fees.
In this context, he also quoted the flawed study ‘only 2% of all VC funding goes to women-led startups’. This study/statistic only considers all-female teams and not mixed teams, and the study was based on data from an opt-in survey.
He briefly mentions that the average VC outperforms the S&P 5001 this is a dumb comparison because:
VCs invest in very early-stage tech and biotech so the correct benchmark is the QQQ + an adjustment for illiquidity and higher risk (see Public Market Equivalent).
IRR itself is a stupid metric because it’s a dollar-weighted metric that is easily manipulated. If you don’t know the difference between dollar- and time-weighted returns, you should not be making any allocations to private investments.
VC is illiquid and its performance measurement continues to suffer from volatility laundering2
(In b4 you say the average/median VC isn’t as relevant because of the Power Law but Silver is comparing the asset class as a whole, and neither you nor I can get an allocation to any of Founders Fund/a16z/Accel’s early stage funds anyway)
There’s an interesting discussion with Patrick Collison regarding the errors of omission and errors of commission where Silicon Valley is biased towards making errors of commission mostly because of FOMO. Interestingly, research on regret shows that errors of omission are less likely to fade over the long-term than errors of commission. But in the short-term the sting from errors of commission hurt more. The reason for this is that errors of commission can be rationalized away “I learned from that”, “I was supporting a friend”, “I was giving back”, etc. but it’s harder to stop fantasizing about what would have happened had the effort/investment succeeded “I could have bought my own private island” because you don’t think about the states of the world where you/your investment failed. So the Regret Minimization Framework is really the Post Hoc Rationalization Framework.
There’s probably some discussion of Annie Duke-style Resulting vs ‘having a good decision-making process’ discussion to be had here, but this post has gone on long enough and I want to start my weekend.
Have a good one!
Silver doesn’t explicitly mention the S&P 500 but he says the average stock market return is 8% so this is the closest approximation. He also uses a simulation of VC fund returns and not actual returns so this comparison is extra dumb.
Volatility laundering is a feature and not a bug if you’re the VC GP.